Sparring over growth and proficiency measures in DeVos’s confirmation hearing

One of the more contentious exchanges during Tuesday night’s confirmation hearing for Betsy DeVos, Donald Trump’s pick for education secretary, began with what seemed like an innocuous, if clumsily worded, request from Sen. Al Franken (D-Minn.). He said to DeVos, “I would like your views on the relative advantage of, measuring, the assessments and using them to measure proficiency or measure growth.”

Franken was referencing debates in education policy about whether test-based accountability for schools should measure students’ proficiency levels or growth in learning over time. DeVos stumbled through a definition of proficiency that sounded more like growth and then asked for clarification, leading Franken to express surprise and exasperation that a nominee for secretary would be so uninformed about a basic, long-debated subject in education.

Proficiency versus growth is, in fact, a long-debated academic question. The question predates 2001’s No Child Left Behind (NCLB) Act, although it was NCLB’s test-based accountability mandates for states that raised its prominence. NCLB initially required all state accountability systems to focus almost exclusively on proficiency rates, setting an aspirational goal that all students would meet proficiency standards by the 2013-14 school year. It provided flexibility for states to create their own standards, assessments, and proficiency thresholds, but required states to hold schools accountable for progressing toward 100 percent proficiency. The proficiency-versus-growth question was also central to then-Secretary Margaret Spellings’s offer to provide waivers to states from NCLB’s accountability mandates back in 2005.

Since that brief time in the limelight, however, this question seems to have faded back into relative obscurity, and now seems to excite only education researchers and accountability hawks. We delve into the issue here, describing what proficiency and growth measures are, why they matter, and what are their relative advantages.

Defining proficiency and growth

The basic proficiency-versus-growth debate revolves around whether schools’ performance on state tests should be evaluated by how their students score relative to an established standard (proficiency) or how much their students’ scores improve over the course of the school year (growth).

Measuring proficiency is straightforward: A student scoring above the proficiency threshold on a particular test is proficient, while a student scoring below the threshold is not. As an approach to evaluating schools, it prioritizes how much students know at the end of a year rather than how much they learned that year.

Growth measures, on the other hand, prioritize how much students learn during a school year rather than how much they know at the end of a year. Based on the same principles as value-added models for teachers, these measures attempt to give credit—or blame—based only on how students’ performance changes as a result of school behaviors. A school that makes substantial progress with an initially low-scoring student population might perform well on growth measures even if many students fall short of proficiency targets.

Reasons to prefer growth to proficiency

There is growing consensus in the education research community that growth measures are generally more appropriate than proficiency measures for evaluating school performance. (Although this sidesteps fierce debate over the role for test-based accountability more broadly.) The principle guiding that view is that schools should only be held accountable for what is realistically in their control. For example, a school that enrolls students who tend to arrive well below grade level seemingly should not be punished for problems caused by prior schools, poverty, or other circumstances.

Franken, in his remarks, touched on another concern emphasizing proficiency: the potential for “bubble effects.” A school that sharply focuses on proficiency may find incentive to direct a disproportionate share of its resources to students it believes are near the proficiency threshold. This form of educational triage might lead schools to regard high-achieving students as “safe” and low-achieving students as “hopeless” with respect to the coming state test, and to allocate its resources accordingly. Researchers have found evidence of disproportionate gains for students on the proficiency bubble—though it’s not entirely clear whether those far below the threshold are left behind.

In spite of this growing consensus about the pitfalls of proficiency, it remains the dominant performance measure in most state accountability systems. Consequently, the proficiency-versus-growth question is still a live issue today. The enactment of the Every Student Succeeds Act (ESSA) in late 2015 threw out many of the more onerous provisions of NCLB, including the 100 percent proficiency target. It did not, however, change the mandate that states employ accountability systems to monitor school performance. And where the text of the law itself was not prescriptive about whether states prioritize proficiency or growth, the new accountability rules put in place under Secretary John King keep the foot on the proficiency pedal (to many education wonks’ chagrin).

Possible roles for proficiency

Even if growth should—though in practice does not—play a more prominent role than proficiency in test-based school accountability, there may be a role for proficiency as well. This results partly from the complicated relationship between growth and proficiency. Consider, for example, a school with already-soaring proficiency rates, whether due to the types of students who enroll or the school’s own success. Is it desirable for this school to devote itself to improving marginally on those state tests—where gains are even possible—or would that be a wasted opportunity to extend and experiment with student learning? It may be that growth on state tests becomes an undesirable pursuit for schools whose students already demonstrate proficiency on state tests.

The growth-versus-proficiency debate also raises intriguing questions about the types of information that states and districts should provide to families choosing schools for their children. In general, a parent might find growth measures helpful in assessing how much her child is likely to learn in a particular school. At the same time, proficiency rates convey information that growth rates cannot. For example, a school’s proficiency rate might reveal information about the overall level of instruction in that school or the knowledge level of a child’s would-be peers. For the reasons discussed above, it might also give a parent context for the school’s growth score. In other words, parents might reasonably incorporate both growth and proficiency measures into their decisionmaking.

Which is correct?

Franken was right that the question of proficiency or growth in test-based accountability is a fundamental question for test-based accountability. It is certainly one that should be of interest to a prospective education secretary—and the education community more broadly—as we continue forward with ESSA implementation and what appears to be a new era of school choice.

While we agree with Franken that there are compelling reasons to prioritize growth measures in school accountability, we question the implication that proficiency versus growth is an either-or proposition.